Recommendation technologies are emerging as an important enabler in helping users identify and navigate their music collections. However, implementing recommendation systems can be expensive in terms of computing resources, and the best recommendation systems rely on a mix of content analysis and social networks to obtain optimal results. As such, resource-constrained devices, such as portable media players, are unable to enjoy the benefits of recommendation technology. Thus, there is a need for a system and method for generating recommendations for a portable media player and then storing, or embedding, the recommendations on the portable media player in such a manner that the recommendations can be accessed at run time with minimal computing resources.